hwihwalab/dama-kurong-solopreneur-ko

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2.6BQuant:BF16Context Size:8kPublished:Aug 24, 2026License:gemmaArchitecture:Transformer Featherless Exclusive Cold

The hwihwalab/dama-kurong-solopreneur-ko is a 2.6 billion parameter, Gemma-2-2B-it based model developed by Hwihwa Lab. It is specifically fine-tuned as an on-device AI companion for solopreneurs, indie hackers, and freelancers, excelling in providing practical bilingual (Korean and English) business advice and support with a consistent fairy companion persona. Optimized for local execution in GGUF format, it incorporates over 50 practical playbooks for tasks like customer support, outage response, and MVP definition.

Loading preview...

Overview

dama-kurong-solopreneur-ko (Kurong) is a lightweight, 2.6 billion parameter AI companion model developed by Hwihwa Lab, fine-tuned from Google's gemma-2-2b-it using Unsloth LoRA. It is designed to act as a supportive desktop companion fairy for solopreneurs, indie hackers, and freelancers, providing practical business advice and assistance.

Key Capabilities

  • Solopreneur Business Playbooks: Incorporates over 50 practical bilingual playbooks (40+ Korean, 10+ English) covering essential tasks like first-paid-user customer support, outage response, MVP pruning, and mental burnout recovery.
  • Dedicated 1st-Person Persona: Maintains a consistent identity as the fairy companion Kurong in both Korean and English conversations.
  • On-Device & GGUF Optimized: Designed for fast, private local inference with a 2B parameter footprint, optimized for platforms like LM Studio, Ollama, and llama.cpp.
  • Bilingual Comprehension: Primarily optimized for natural, polite Korean dialogue and localized business advice, while retaining Gemma-2's multilingual and English instruction-following capabilities.

Good For

  • Solopreneurs and Freelancers: Seeking an AI assistant for business-related advice and operational support.
  • Korean-speaking Users: Optimized for natural and polite Korean conversational tone.
  • Local Inference: Users who prefer running models on their own devices for privacy and speed, utilizing GGUF formats.
  • Interactive Companions: Those interested in an AI with a distinct, supportive persona for daily interactions.

Finetunes
1 models